MLOps Coding Skills

Welcome to the MLOps Coding Skills repository! 🚀

This project contains a collection of Agent Skills designed to accompany the MLOps Coding Course. These skills are tailored to help AI coding assistants (like Gemini, Copilot, and Claude) understand and guide you through the best practices of MLOps, from prototyping to production.

🧠 What are Agent Skills?

Agent Skills are specialized context or instructions that you can load into your AI assistant. They provide the "Know-How" for specific tasks, ensuring that the AI follows the course's methodology, standards, and patterns.

This repository covers skills for:

• Initialization: Setting up robust project structures.
• Prototyping: Writing clean, experimental code.
• Industrialization: Optimizing for production deployment.
• Validation: Testing data, models, and code.
• Automation: CI/CD and workflow automation.
• Collaboration: Working effectively in teams.
• Observability: Implementing logging, lineage, and monitoring.

🛠️ Installation & Usage

Recommended Setup (Symlinks)

To use these skills alongside others, we recommend creating a .agent/skills folder in your project directory and creating symlinks to the specific skills you need from this repository.

1. Clone this repository to a location on your machine (e.g., ~/mlops-coding-skills).

2. Create a skills directory in your target project or workspace:

   mkdir -p .agent/skills

3. Symlink the skills you want to use. For example, to use all MLOps skills:

   # Example: Symlink all skill folders
   ln -s ~/mlops-coding-skills/mlops-* .agent/skills/

1. Antigravity

Antigravity natively supports the .agent/skills standard. If you have set up the symlinks as described in the Recommended Setup section, Antigravity will automatically detect and use these skills.

See Antigravity Skills Docs for more details.

2. Gemini CLI

The Gemini CLI manages skills in .gemini/skills. We recommend installing this repository as a remote skill set to keep it updated and scoped correctly.

Install via Git (Recommended) This command installs the skills into your current workspace scope.

gemini skill install https://github.com/MLOps-Courses/mlops-coding-skills --scope=workspace

See Gemini CLI Skills Docs for more details.

3. GitHub Copilot

For GitHub Copilot, you can provide these skills as context in your workspace.

Agent Mode: If you are using Copilot in Agent Mode (e.g., in VS Code), you can explicitly reference the skill files when asking for help with a specific MLOps task.

1. Open your project in VS Code.
2. Open the relevant SKILL.md file (e.g., .agent/skills/mlops-observability/SKILL.md) to load it into context.
3. Ask Copilot: "Follow the instructions in the open skill file to help me implement observability."

See GitHub Copilot Agent Skills for the latest features.

4. Claude Desktop / Claude Code

For Claude, you can use these skills as "Project Knowledge" or explicit context.

Claude Desktop:

1. Create a "Project" in Claude.
2. Add the relevant SKILL.md files from this repository (or your .agent/skills folder) to the Project Knowledge.
3. Claude will now follow these guidelines when answering your queries within that project.

See Claude Skills Documentation for more advanced integrations.

🤝 Contributing

We welcome contributions! If you have a new MLOps pattern or improvement, please open a Pull Request.

Happy Coding! 🤖✨